|Publication number||US7383504 B1|
|Application number||US 09/518,937|
|Publication date||Jun 3, 2008|
|Filing date||Mar 6, 2000|
|Priority date||Aug 30, 1999|
|Publication number||09518937, 518937, US 7383504 B1, US 7383504B1, US-B1-7383504, US7383504 B1, US7383504B1|
|Inventors||Ajay Divakaran, Anthony Vetro, Huifang Sun|
|Original Assignee||Mitsubishi Electric Research Laboratories|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (30), Non-Patent Citations (31), Referenced by (25), Classifications (10), Legal Events (6)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This is a Continuation-in Part application of U.S. patent application Ser. No. 09/385,169, “Method for Representing and Comparing Multimedia Content” filed on Aug. 30, 1999 now U.S. Pat. No. 6,546,135 by Lin et.
This invention relates generally to processing multimedia content, and more particularly, to representing and comparing ranked multimedia content.
There exist many standards for encoding and decoding multimedia content. The content can include audio signals in one dimension, images with two dimensions in space, video sequences with a third dimension in time, text, or combinations thereof. Numerous standards exist for audio and text.
For images, the best known standard is JPEG, and for video sequences, the most widely used standards include MPEG-1, MPEG-2 and H.263. These standards are relatively low-level specifications that primarily deal with the spatial compression in the case of images, and spatial and temporal compression for video sequences. As a common feature, these standards perform compression on a frame basis. With these standards, one can achieve high compression ratios for a wide range of applications.
Newer video coding standards, such as MPEG-4, see “Information Technology—Generic coding of audio/visual objects,” ISO/IEC FDIS 14496-2 (MPEG4 Visual), November 1998, allow arbitrary-shaped objects to be encoded and decoded as separate video object planes (VOP). This emerging standard is intended to enable multimedia applications, such as interactive video, where natural and synthetic materials are integrated, and where access is universal. For example, one might want to “cut-and-paste” a moving figure or object from one video to another. In this type of scenario, it is assumed that the objects in the multimedia content have been identified through some type of segmentation algorithm, see for example, U.S. patent application Ser. No. 09/326,750 “Method for Ordering Image Spaces to Search for Object Surfaces” filed on Jun. 4, 1999 by Lin et al.
The most recent standardization effort taken on by the MPEG committee is that of MPEG-7, formally called “Multimedia Content Description Interface,” see “MPEG-7 Context, Objectives and Technical Roadmap,” ISO/IEC N2729, March 1999. Essentially, this standard plans to incorporate a set of descriptors and description schemes that can be used to describe various types of multimedia content. The descriptor and description schemes are associated with the content itself and allow for fast and efficient searching of material that is of interest to a particular user. It is important to note that this standard is not meant to replace previous coding standards. Rather, it builds on other standard representations, especially MPEG-4, because the multimedia content can be decomposed into different objects and each object can be assigned a unique set of descriptors. Also, the standard is independent of the format in which the content is stored. MPEG-7 descriptors can be attached to compressed or uncompressed data.
Descriptors for multimedia content can be used in a number of ways, see for example “MPEG-7 Applications,” ISO/IEC N2728, March 1999. Most interesting, for the purpose of the description below, are database search and retrieval applications. In a simple application environment, a user may specify some attributes of a particular object. At this low-level of representation, these attributes may include descriptors that describe the texture, motion and shape of the particular object. A method of representing and comparing shapes has been described in U.S. patent application Ser. No. 09/326,759 “Method for Ordering Image Spaces to Represent Object Shapes” filed on Jun. 4, 1999 by Lin et al. One of the drawbacks of this type of descriptor is that it is not straightforward to effectively combine this feature of the object with other low-level features. Another problem with such low-level descriptors, in general, is that a high-level interpretation of the object or multimedia content is difficult to obtain. Hence, there is a limitation in the level of representation.
To overcome the drawbacks mentioned above and obtain a higher-level of representation, one may consider more elaborate description schemes that combine several low-level descriptors. In fact, these description schemes may even contain other description schemes, see “MPEG-7 Description Schemes (V0.5),” ISO/IEC N2844, July 1999.
As shown in
The major problem with such a scheme is that the relations and attributes specified by the syntactic and semantic DS are independent, and it is the burden of the relation graph DS to create a coherent and meaningful interpretation of the multimedia content. Furthermore, the DSs mentioned above are either tree-based or graph-based. Tree-based representations provide an efficient means of searching and comparing, but are limited in their expressive ability; the independent syntactic and semantic DSs are tree-based. In contrast, graph-based representations provide a great deal of expressive ability, but are notoriously complex and prone to error for search and comparison.
For the task at hand, it is crucial that a representation scheme is not limited to how multimedia content is interpreted. The scheme should also provide an efficient means of comparison. From a human perspective, it is possible to interpret multimedia content in many ways; therefore, it is essential that any representation scheme allows multiple interpretations of the multimedia content. Although the independent syntactic and semantic DS, in conjunction with the relation graph DS, may allow multiple interpretations of multimedia content, it would not be efficient to perform comparisons.
As stated above, it is possible for a DS to contain other DSs. In the same way that the generic DS includes a syntactic DS, a semantic DS, and a syntactic/semantic relation graph DS. It has been proposed that the syntactic DS 101 includes a segment DS 105, a region DS 106, and a segment/region relation graph DS 107. As shown in
Therefore, there is a need for representing syntactic and semantic attributes of multimedia content that balances the complexities of data structures and the methods that operate on the structures. In addition, there is a need to compare multimedia content according to content attributes.
The present invention provides a new method of representing syntactic and semantic attributes of multimedia content. It is an object of the invention to use existing attributes that may be contained within a semantic or syntactic description scheme using a framework that balances the restrictions on structure and expressiveness of elements, with the computational complexity of operations on those elements.
The method according to the invention is based in part on directed acyclic graphs (DAG). It is well known that the DAG occupies a middle ground between tree-based and graph-based representations. In addition, the DAG provides a new functionality of composition. In other words, many structural compositions of an entity can be described by many structural compositions of its contained elements.
Most importantly though, the similarity between these structural compositions and the structural compositions created by another entity can be easily computed. Within this composition framework, the DAG also provides a means of combining syntactic and semantic elements so that similarity comparisons may seamlessly switch between both types of descriptions. In some sense, this can be viewed as a unification between the syntactic and semantic parts of the description scheme.
The method for representing the semantic and syntactic elements in a unified way also provides a means for unifying the spatial and temporal elements of multimedia content. The invention relies on the fact that the compositions referred to earlier are spatio-temporal compositions that contain both syntactic and semantic elements. The important points to keep in mind is that the compositions according to the invention are DAG representations, which facilitate multiple interpretations and low complexity comparison, and that the compositions, which define spatio-temporal attributes, both syntactic and semantic, are contained within the respective content entities.
More particularly, the method generates a representation of multimedia content by first segmenting the multimedia content spatially and temporally to extract objects. Feature extraction is applied to the objects to produce semantic and syntactic attributes, relations, and a containment set of content entities. The content entities are coded to produce directed acyclic graphs of the content entities. Edges of the directed acyclic graphs represent the content entities, and nodes represent breaks in the segmentation. Each directed acyclic graph represents a particular interpretation of the multimedia content.
In one aspect the multimedia content is a two dimensional image, and in another aspect the multimedia content is a three dimensional video sequence.
In a further aspect of the invention, representations for different multimedia contents are compared based on similarity scores obtained for the directed acyclic graphs. Attributes of each content entity are measured and the measured attributes are assigned to each corresponding content entity in the directed acyclic graphs to rank order the multimedia content.
In another aspect of the invention, attributes of each content entity are measured, and the entities are ranked according to the measured attributes. The rank list can be culled for desirable permutations of primary content entities as well as secondary entities associated with the primary entities. By culling desirable permutations, one can summarize, browse or traverse the multimedia content. For example, the most active and least active video segments of a video sequence form a summary that has the desirable attribute of conveying the dynamic range of action contained in the video sequence.
We describe methods for representing and comparing multimedia content according to a ranking of the content. The methods are based a new generic data structure, which includes a directed acyclic graph (DAG) representation. In the following, we describe objects in our scheme and the advantages of the DAG representation. It is the DAG representation that allows the scheme to infer multiple interpretations of multimedia content, yet still be efficient in the comparison with other multimedia content. In fact, when we score with respect to a probability likelihood function, the computations are not only tractable, but also optimal.
Besides describing the generic data structure, we also describe three important functions that allow us to realize this efficient representation and perform comparisons. The first function will be referred to as a DAG-Coder. The DAG-Coder is responsible for taking individual content entities contained in the object and producing a DAG-Composition. The second function is an Object-Compare. The Object-Compare efficiently compares two content entities by determining a similarity score. The third function is Content Ranker. This function ascribes a ranking score to content entities so that DAG-Compositions can be traversed, browsed, or summarized according to rank. The traversing, browsing, and summarizing can be an increasing or decreasing rank order.
After the data structure and three functions mentioned above have been described, we review and elaborate on applications that are enabled by our representation scheme. An integrated application system that performs feature extraction, database management and object comparison is described. Also described is an application system for traversing, browsing, and summarizing multimedia content according to a ranking of the content.
Generic Description Scheme of a Content Entity
To introduce our scheme of representing content objects, we define generic object types, and restrictions on instantiations of such generic object types.
As shown in
The attributes 201 form the basis within our recursive description scheme. Attributes are an unordered set that contains properties that may provide details about parts of the entity or summarize the entity as a whole. Attributes are global to the object and may refer to such syntactic properties as color and motion, or other semantic properties of the object such as time and place. The attributes provide basic, low-level information without any structure, however, after structure is added, it is these properties that actually contribute to the degree of similarity. Also, as we will describe later, attributes can define an ordering that help to compose and interpret the individual entities contained within the object. It should be noted that these properties are inherent qualities of the content entity that contains them and instantiations of this entity should be accessible/visible through the content entity itself.
As an example, a video sequence of a airplane landing on a runway may contain the semantic attributes of place, date, time and temperature, along with the caption, “airplane (767) landing.” Some syntactic attributes that may be attached to this multimedia content are the trajectory of descent. Attached to the airplane object may be the color and shape of the airplane itself. Here, we make an important distinction between attributes of the multimedia content and attributes of the objects. The reason that the trajectory is an attribute of the multimedia content is because trajectory is relative to the ground. Therefore, it does not make sense as an attribute of the plane alone, whereas color and shape do make sense.
The relations (R) 202 are objects that detail relationships between content entities (VE). It is important to note that the context of the relations is given by the containing content entity. The reason is that multimedia content that are segmented differently will produce different relations. Essentially, the relation can be viewed as a hyperlink between a contained object and something else, for example, another content entity. Types of relations are global and instantiations of relations should only be accessible within the content entity itself. One of the utilities of relations is that they may be useful in guiding a search. Returning to our example of the airplane landing, several relations can be identified: the plane is landing on the runway, the lights are guiding the plane, and the runway is located at a particular airport with a particular orientation.
The relations are different from containment, described below, in that the related object may not be completely contained by the content entity and therefore is not considered in similarity comparisons. However, relations allow a user to search for other relevant objects to the content entity in question. All the relations in the content entity must have one argument that is contained within the content entity.
In general, the DAG-Compositions 203 are directed acyclic graphs 205 where edges 206 represent content entities and nodes 207 correspond to breakpoints in the segmentation. The DAG-Composition allow us to infer multiple interpretations of the same multimedia content. Because DAGs operate on 1D spaces, segmentation in this context refers to the delineation of some 1D process. For instance, if we consider a spatio-temporal multimedia content, then the temporal segmentation is a 1D process that defines points in time where several successive events may begin and end. Hence, we may have a DAG-Composition that corresponds to temporal actions. In the spatial domain, we may define an order from left to right across an image. In this way, we may have a DAG-Composition that corresponds to object positions from left to right. Of course, we may define other orderings such as a counter-clockwise spatial ordering, which may serve a totally different purpose.
In U.S. patent application Ser. Nos. 09/326,750 and 09/326,759, incorporated herein by reference, Voronoi ordering functions were respectively defined over the exterior and interior image space with respect to an object boundary. The ordering on the interior space was particularly useful in obtaining a skeleton-like representation of the object shape, then forming a partially ordered tree (POT), which made use of the DAG representation.
It should be emphasized though that the method of ordering 2D images or 3D video sequences to achieve DAG-Compositions is not the focus here, rather we are concerned with techniques that use the DAG-Composition to infer higher-level interpretations of a particular multimedia content.
The containment set 204 includes pointers to other content entities that are strictly contained temporally and/or spatially within the content entity 200. The restriction on the containment set is that one object cannot contain another object that contains the first object, i.e., containment induces a directed acyclic graph. The content entities need not be mutually exclusive and there is no ordering within the containment set. For example, in the video sequence of the airplane landing, the containment set includes pointers to each content entity. Some possibilities include pointers to the plane, the runway, the runway lights, the plane touching down, radio communications, etc.
The DAG-Compositions are the result of different DAG-Coders applied to the content entity. In other words, given the content entities in the containment set and their relations, different DAG-Coders produce different interpretations of the multimedia content. This function is further described in the following.
A DAG-Coder is a function that segments a given content entity into its components by inducing an ordering over the content entity components. The DAG-Coder produces the DAG-Composition 204. The DAG-Coder is global to the database and can be applied to any content entity. The DAG-Coder provides a perspective on the spatio-temporal content space and make similarity calculations between objects more tractable. A path in the DAG represents an interpretation of the content entity 200. This DAG representation becomes a framework for the description scheme that can interchange syntactic and semantic information, at any level. Furthermore, the complexity of the description scheme is hidden from the user.
Multiple Path Through a DAG
The DAG-Coder produces multiple interpretations of the multimedia content through such DAG-Compositions. This is achieved through the multiple path structure of the DAG. In the following, we focus on what these multiple paths really mean in terms of the multimedia content.
As illustrated, a temporal DAG can represent equivalent interpretations of the same event. For instance as shown in
In the case of spatial DAGs, multiple paths can also represent equivalent interpretations, and in some sense can add a higher level of expressiveness. This added level of expressiveness is achieved by a grouping of individual objects into a composite object, then realizing that this composite object can be interpreted with a different semantic meaning. Usually, this new semantic interpretation is higher than before since more information is considered as a whole.
As an example, consider several objects: a gasoline pump, a gas attendant and a car. Individually, these objects have their own set of attributes and are distinct in their semantic meaning. Put together though, these individual objects can obviously be interpreted as a gas station. These multiple paths are efficiently represented by the DAG structure. On the syntactic side, various interpretations of the shape of an object for example may be deduced in a similar manner.
Generating Multimedia Content Description
Comparing Different Multimedia Content
We consider the matching algorithms used to compare Ordered Trees. Because trees are recursive structures, we can do optimal comparisons recursively and base the comparisons upon single node trees. Let us score our trees in the range of 0 to 1. Two single node trees are assigned a score of 1, while a single node tree and any tree with more than one node is assigned a score of 0.
For our inductive step, we note that each node corresponds to a sequence of edges and their respective children. To compare trees, we merely find the best correspondence between the sequences, while recursively comparing their corresponding children. A Largest Common Subsequence (LCS) matching can be used for this step. The Object-Compare methods allows efficient, robust and optimal comparison of objects at the same complexity of the Ordered Tree comparisons.
To handle the extra freedom in the expressiveness of DAGs, we use a DAG-Compare algorithm, see “Lin, et al., “Coding and Comparison of DAGs as a novel neural structure with application to on-line handwritten recognition,” IEEE Trans Signal Processing, 1996, incorporated herein by reference. We find the two best-matching paths between two DAGs. Although more general, the DAG-Compare is of the same order complexity as the LCS search. Lastly, we should mention that the constraints on the containment hierarchy (as a DAG) allow us to use the POT-Compare algorithm, but the POT is merely a subset of our generic content entity.
The content description scheme described above is not only an expressive means of describing content entities, but also provides a robust similarity measure that is computationally efficient and can seamlessly integrate various descriptions, both semantic and syntactic. Within the description scheme according to our invention, content entities, their attributes and their relations form a basic hyperlink network such as available from the HTTP standard.
By constraining our graph structures to Directed Acyclic Graphs onto the hierarchy of our content entities and their descriptions, we can give an extra expressiveness over ordered trees while maintaining computational complexity for robust comparison between content entities that is equivalent to an ordered tree comparison.
Freedom in Expressiveness
There is no strict hierarchy of content entities: any object may strictly contain another object as long as the containment is not contradictory. Instead of a tree hierarchy, the containment relation over the content entities induces a directed acyclic graph. Acircularity is maintained by disallowing contradictory containment. The restriction on cycles enables an efficient recursive formulation of comparison.
Focusing on the DAG structure, we map the DAG structure of DAG-Composition as follows: edges represent content entities, and nodes correspond to breakpoints segmentation. We can structure the object as a configuration of contained content entities within DAGs according to a predefined topological order. The restrictions on the DAGs compared to a general graph structure is its topological ordering. This order may be temporal or spatial, but it must be 1D. By following the order and obeying connectivity, a subgraph of DAG structure leads to a new concept: an ordered path represents a particular interpretation of multimedia content, i.e. a representative view of the content entity as an ordered subset of its contained entities.
Because a DAG can contain multiple ordered paths, the DAG becomes a compact representation of the multiple interpretations of the data. The DAG data structure allows for the concept of parallel paths; thus, the DAG may integrate both semantic and syntactic elements through this parallel structure. The semantic and syntactic elements are not necessary equivalent, but, within the context of the DAG structure, they can be made interchangeable by placing them on these parallel constructions and its ordering.
These functionalities are a subset of a generic graph structure. However, as most graph matching problems are still open, these restrictions will allow us to compare these expressive structures. Although this ordering constrains the expressiveness of a DAG-composition, it does allow for element alignment in robust comparison of content entities.
Universal Multimedia Access
Because our description scheme is capable of representing and comparing multiple interpretations of multimedia content, it fits very well with the concept of Universal Multimedia Access (UMA). The basic idea of UMA, as shown in
Recently, several solutions have focussed on adapting the multimedia content to the client devices. UMA can be provided in two basic ways—the first by storing, managing, selecting, and delivering different versions of the media objects (images, video, audio, graphics and text) that comprise the multimedia presentations. The second way is by manipulating the media objects on-the-fly, such as by using methods for text-to-speech translation, image and video transcoding, media conversion and summarization. This allows the multimedia content delivery to adapt to the wide diversity of client device capabilities in communication, processing, storage, and display.
Our description scheme can support UMA through the first item mentioned above, that is, depending on the client-side capabilities, the server-side may choose to send a more elaborate interpretation of the multimedia content or simply send a brief summary of the multimedia content. In this way, our description scheme acts as a managing structure that helps decide which interpretation of the multimedia content is best suited for the client-side devices. As part of the attributes for a content entity, the requirements may include items such as the size of each image or video frame, the number of video frames in the multimedia content, and other fields that pertain to resource requirements.
As an additional feature, the content entities in a DAG can have associated ranks.
For example, the various segments of an “adventure-action” movie video can be ranked on a scale of 1-10 as to the intensity of the “action” in the movie. Similarly, the segments of a sports video, such as a football match can be ranked, where a scoring opportunity receives a relatively high score, and an “injury” on the field receives a relatively low score. Segments of gothic romance videos can be ranked on the relative level of “romantic” activity, horror films on the level of the levels of fright inducing scenes, comedies on their level of humor, rock videos on their loudness, and so forth. It should be understood that the measurements can be based on the semantic and/or the syntactic properties of the content.
The ranking can be manual, or machine generated. For example, a high number of short segments in a row would generally be indicative of a high level of activity, whereas long segments would tend to include a low level of activity. See Yeo et al., in “Rapid Scene Analysis on Compressed Video,” IEEE Transactions on Circuits and Systems for Video Technology, Vol. 5, No. 6, December 1995, page 533-544, for one way of measuring content attributes.
Once the various segments have been ranked, as shown in
Specifying a particular rank-based traversal in effect allows one to summarize a video. The “summary” shown in
In another embodiment, as shown in
Although the invention has been described by way of examples of preferred embodiments, it is to be understood that various other adaptations and modifications may be made within the spirit and scope of the invention. Therefore, it is the object of the appended claims to cover all such variations and modifications as come within the true spirit and scope of the invention.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4953106 *||May 23, 1989||Aug 28, 1990||At&T Bell Laboratories||Technique for drawing directed graphs|
|US5297150 *||Jun 17, 1992||Mar 22, 1994||International Business Machines Corporation||Rule-based method for testing of programming segments|
|US5412774 *||Nov 3, 1993||May 2, 1995||At&T Corp.||Apparatus for and method of displaying a data item of a database using the display function of a selected data item|
|US5487117 *||Oct 11, 1994||Jan 23, 1996||At&T Corp||Graphical system for automated segmentation and recognition for image recognition systems|
|US5511158 *||Aug 4, 1994||Apr 23, 1996||Thinking Machines Corporation||System and method for creating and evolving directed graphs|
|US5521841 *||Mar 31, 1994||May 28, 1996||Siemens Corporate Research, Inc.||Browsing contents of a given video sequence|
|US5673369 *||Mar 2, 1995||Sep 30, 1997||International Business Machines Corporation||Authoring knowledge-based systems using interactive directed graphs|
|US5708767 *||Feb 3, 1995||Jan 13, 1998||The Trustees Of Princeton University||Method and apparatus for video browsing based on content and structure|
|US5790416 *||Sep 18, 1995||Aug 4, 1998||Motorola, Inc.||Updating hierarchical DAG representations through a bottom up method|
|US5821945||May 15, 1997||Oct 13, 1998||The Trustees Of Princeton University||Method and apparatus for video browsing based on content and structure|
|US5822537 *||Apr 5, 1996||Oct 13, 1998||At&T Corp.||Multimedia networked system detecting congestion by monitoring buffers' threshold and compensating by reducing video transmittal rate then reducing audio playback rate|
|US5828809 *||Oct 1, 1996||Oct 27, 1998||Matsushita Electric Industrial Co., Ltd.||Method and apparatus for extracting indexing information from digital video data|
|US5838830 *||Apr 3, 1997||Nov 17, 1998||Sharp Laboratories Of America, Inc.||Vertex-based hierarchical shape representation and coding method and apparatus|
|US5873081 *||Jun 27, 1997||Feb 16, 1999||Microsoft Corporation||Document filtering via directed acyclic graphs|
|US5956026 *||Dec 19, 1997||Sep 21, 1999||Sharp Laboratories Of America, Inc.||Method for hierarchical summarization and browsing of digital video|
|US6002803 *||Aug 19, 1997||Dec 14, 1999||Sharp Laboratories Of America, Inc.||Methods of coding the order information for multiple-layer vertices|
|US6049332 *||Oct 7, 1996||Apr 11, 2000||Sony Corporation||Method and apparatus for the scheduling and ordering of elements in a multimedia environment|
|US6079566 *||Apr 7, 1998||Jun 27, 2000||At&T Corp||System and method for processing object-based audiovisual information|
|US6108448 *||Jun 12, 1997||Aug 22, 2000||International Business Machines Corporation||System and method for extracting spatially reduced image sequences in a motion compensated compressed format|
|US6236395 *||Apr 26, 1999||May 22, 2001||Sharp Laboratories Of America, Inc.||Audiovisual information management system|
|US6243419 *||May 27, 1997||Jun 5, 2001||Nippon Telegraph And Telephone Corporation||Scheme for detecting captions in coded video data without decoding coded video data|
|US6266053 *||Apr 3, 1998||Jul 24, 2001||Synapix, Inc.||Time inheritance scene graph for representation of media content|
|US6292938 *||Dec 2, 1998||Sep 18, 2001||International Business Machines Corporation||Retargeting optimized code by matching tree patterns in directed acyclic graphs|
|US6356902 *||Jul 28, 1998||Mar 12, 2002||Matsushita Electric Industrial Co., Ltd.||Method and system for storage and retrieval of multimedia objects|
|US6516090 *||Apr 23, 1999||Feb 4, 2003||Canon Kabushiki Kaisha||Automated video interpretation system|
|US6618444 *||Dec 21, 1999||Sep 9, 2003||At&T Corp.||Scene description nodes to support improved chroma-key shape representation of coded arbitrary images and video objects|
|US6618507 *||Jan 25, 1999||Sep 9, 2003||Mitsubishi Electric Research Laboratories, Inc||Methods of feature extraction of video sequences|
|US6665732 *||Aug 21, 1998||Dec 16, 2003||Lucent Technologies Inc.||Method and system for resource scheduling composite multimedia objects|
|US20020159640 *||May 6, 2002||Oct 31, 2002||Philips Electronics North America Corporation||Meta-descriptor for multimedia information|
|WO1999041684A1 *||Feb 11, 1999||Aug 19, 1999||Fast Tv||Processing and delivery of audio-video information|
|1||*||A. Puri, et al., "MPEG-4: An Object-based multimedia coding standard support mobile application," Mobile Networks and Application, vol. 3, Issue 1, ACM, Jun. 1998, pp. 5-32.|
|2||*||A. Vetro, et al., "MPEG-4 Rate Control for Multiple Video Objects," IEEE, Feb. 1999, pp. 186-199.|
|3||*||Alatan et al., Image Sequence Analysis for Emerging Interactive Multimedia Services-The European COST 211 Framework,IEEE Transactions on Circuits and Systems for Video Technology, vol. 8, No. 7, Nov. 1998.|
|4||*||B. Gunsel, et al., "Temporal Video Segmentation Using Unsupervised Clustering and Semantic Object Tracking," Journal of Electronic Imaging, Jul. 1998, pp. 592-604.|
|5||Chang, S.-F. et al. "VideoQ: an automated content based video search system using visual cues" Proceedings ACM Multimedia, 1997. vol. 5. Nov. 9-13, 1997, pp. 313-324.|
|6||Day, Y.F. et al. "Object-Oriented Conceptual Modeling of Video Data" Proceedings of the 11<SUP>th </SUP>Annual International Conference on Data Engineering, Taipei, Taiwan Mar. 6-10, 1995, pp. 401-408.|
|7||*||Divakaran et al., "Report on Validation Experiment on Ordered Relation Graphs", Dec. 1999, pp. 1-14.|
|8||*||E. Wold et al., "Content-Based Classification, Search, and Retrieval of Audio," IEEE, Fall 1999, pp. 27-36.|
|9||*||Friedman et al., "Image Segmentation in video sequences: A probabilistic approach", Computer Science Division, 1997, pp. 1-13.|
|10||*||G. Huhne, et al., "Transmitting MPEG-4 Video Streams over the Internet: Problems and Solutions," ACM, Oct. 1999, pp. 135-138.|
|11||*||Gene Miller et al., "News On-Demand For Multimedia Networks", ACM, 1993, pp. 1-11.|
|12||*||I. Lin, et al., "Automatic Video Object Segmentation Via Voronoi Ordering and Surface Optimization," IEEE, 1999, pp. 265-270.|
|13||*||IBM Technical Disclosure Bulletin, "Encoding Data Into Irrational Magic Numbers for Fast Searching and Comparing", Nov. 1993, Volume No. 36, pp. 1-2.|
|14||*||I-Jong Lin et al., "Efficient Representation and Comparison of Multimedia Content Using Dag-Composition", Princeton University, Department of Electrical Engineering, 2000, pp. 895-898.|
|15||*||J.M. Gauch et al., "Real Time Video Scene Detection and Classification," Information Processing and Management, May 1999, pp. 381-400.|
|16||Lau, R.W.H. et al. "VideoMAP: a generic framework for video management and application processing" Proceedings of the 33<SUP>rd </SUP>Annual Hawaii International Conference on System Sciences, 2000. Jan. 4, 2000, pp. 3218-3227.|
|17||*||Meier et al., "Video Segmentation for Content-Based Coding", IEEE Transactions On Circuits and System for Video Technology, vol. 9, No. 8, pp. 1190-1203,Dec. 1999.|
|18||*||N. Brady, "MPEG-4 Standardized Methods for the Compression of Arbitrarily Shaped Video Objects," IEEE, Dec. 1999, pp. 1170-1189.|
|19||*||R. Lienhart et al., "Scene Determination Based on Video and Audio Features," IEEE, Jun. 7-11, 1999, pp. 685-690.|
|20||*||S. Das, et al., "Self-Stabilizing Algorithm in DAG Structured Networks," IEEE, Jun. 1999, pp. 190-195.|
|21||*||Salembier et al., "Region-Based Representations of Image and Video: Segmentation Tools for Multimedia Services", IEEE Transactions on Circuits and Systems for Video Technology, vol. 9, No. 8, Dec. 1999.|
|22||*||T. Zhang et al., "Hierarchical Classification of Audio Data for Archiving and Retrieving," IEEE, Mar. 15-19, 1999, pp. 3001-3004.|
|23||*||Taekyong Lee et al., "Queryin Processing Techniques for Multimedia Presentation Graphs", Department of Computer Engineering and Science, 1998, pp. 130-138.|
|24||*||Taekyong Lee et al., "Querying Multimedia Presentations Based on Content", IEEE Transactions On Knowledge and Data Engineering, vol. 3, May/Jun. 1999, pp. 361-385.|
|25||*||Trier, et al., "Feature Extraction Methods for Character Recognition- A Survey," 1996, pp. 641-662.|
|26||*||Y. He, et al., "A Software-Based MPEG-4 Video Encoder Using Parallel Processing," IEEE, Nov. 1998, pp. 909-920.|
|27||*||Y. He, et al., "Dynamic Scheduling of Multiple Video Objects for MPEG-4 Encoding with User Interactions," IEEE, 1999, pp. 319-322.|
|28||*||Y. Wang et al., "Multimedia Content Classification Using Motion and Audio Information," IEEE, Jun. 9-12, 1997, Hong Kong, pp. 1488-1491.|
|29||*||Yeo et al., "A Unified Approach to Temporal Segmentation of Motion JPEG and MPEG Compressed Video", IEEE, pp. 81-88, May 1995.|
|30||*||Yeo et al., "Rapid Scene Analysis on Compressed Video", IEEE Transaction On Circuits and System for Video Technology, vol. 5, No. 6, pp. 533-544, Dec. 1995.|
|31||*||Yeo et al., "Time-constrained Clustering for Segmentation of Video into Story Units", IEEE Proceedings of ICPR'96, pp. 375-380, 1996.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7624346 *||Jul 26, 2002||Nov 24, 2009||National Institute Of Information And Communications Technology Incorporated Administrative Agency||Contents summarizing unit and contents summarizing program that produce content summary information summarized from contents regarding sports|
|US7765322 *||Aug 2, 2005||Jul 27, 2010||International Business Machines, Corporation||System for executing a multimedia resource|
|US7908254 *||Jun 10, 2005||Mar 15, 2011||Hewlett-Packard Development Company, L.P.||Identifying characteristics in sets of organized items|
|US8024356 *||Sep 20, 2011||Autodesk, Inc.||Database-managed image processing|
|US8385672 *||Apr 30, 2008||Feb 26, 2013||Pictometry International Corp.||System for detecting image abnormalities|
|US8614718 *||May 14, 2013||Dec 24, 2013||Disney Enterprises, Inc.||Method and system for producing a rich multimedia interaction|
|US8645403||Feb 3, 2006||Feb 4, 2014||Autodesk, Inc.||Database-managed rendering|
|US8661096 *||Nov 5, 2007||Feb 25, 2014||Cyberlink Corp.||Collaborative editing in a video editing system|
|US9208218 *||Oct 19, 2011||Dec 8, 2015||Zalag Corporation||Methods and apparatuses for generating search expressions from content, for applying search expressions to content collections, and/or for analyzing corresponding search results|
|US9244924 *||Jan 9, 2013||Jan 26, 2016||Sri International||Classification, search, and retrieval of complex video events|
|US9262818||Jan 17, 2013||Feb 16, 2016||Pictometry International Corp.||System for detecting image abnormalities|
|US20040236805 *||May 23, 2003||Nov 25, 2004||Goren Gordon||Method and system for creating meaningful summaries from interrelated sets of information units|
|US20050262141 *||Jul 26, 2002||Nov 24, 2005||National Institute Of Onformation And Communicatio Ns Technology Incorporated Administrative Agency||Content summarizing apparatus and content summarizing program|
|US20060077992 *||Aug 2, 2005||Apr 13, 2006||Gale Martin J||System for executing a multimedia resource|
|US20060282475 *||Jun 10, 2005||Dec 14, 2006||Suermondt Henri J||Identifying characteristics in sets of organized items|
|US20070185881 *||Feb 3, 2006||Aug 9, 2007||Autodesk Canada Co.,||Database-managed image processing|
|US20070198584 *||Feb 3, 2006||Aug 23, 2007||Autodesk Canada Co.||Database-managed rendering|
|US20080273753 *||Apr 30, 2008||Nov 6, 2008||Frank Giuffrida||System for Detecting Image Abnormalities|
|US20090119369 *||Nov 5, 2007||May 7, 2009||Cyberlink Corp.||Collaborative editing in a video editing system|
|US20120054635 *||Aug 8, 2011||Mar 1, 2012||Pantech Co., Ltd.||Terminal device to store object and attribute information and method therefor|
|US20120102469 *||Apr 26, 2012||International Business Machines Corporation||Deterministic application breakpoint halting by logically relating breakpoints in a graph|
|US20120330943 *||Sep 5, 2012||Dec 27, 2012||Thomson Licensing S.A.||Simplified searching for media services using a control device|
|US20130103662 *||Oct 19, 2011||Apr 25, 2013||Zalag Corporation||Methods and apparatuses for generating search expressions from content, for applying search expressions to content collections, and/or for analyzing corresponding search results|
|US20130293562 *||May 14, 2013||Nov 7, 2013||Disney Enterprises, Inc.||Method and System for Producing a Rich Multimedia Interaction|
|US20160034786 *||Jul 29, 2014||Feb 4, 2016||Microsoft Corporation||Computerized machine learning of interesting video sections|
|U.S. Classification||715/723, 382/190, 345/440, 382/232|
|International Classification||G06T11/20, G06F3/00|
|Cooperative Classification||G06K9/00711, G06K9/6878|
|European Classification||G06K9/68C, G06K9/00V3|
|Mar 6, 2000||AS||Assignment|
Owner name: MITSUBISHI ELECTRIC INFORMATION TECHNOLOGY CENTER
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:DIVAKARAN, AJAY;VETRO, ANTHONY;SUN, HUIFANG;REEL/FRAME:010657/0822
Effective date: 20000303
|Jan 23, 2001||AS||Assignment|
Owner name: MITSUBISHI ELECTRIC RESEARCH LABORATORIES, INC., M
Free format text: CHANGE OF NAME;ASSIGNOR:MITSUBISHI ELECTRIC INFORMATION TECHNOLOGY CENTER AMERICA, INC.;REEL/FRAME:011564/0329
Effective date: 20000828
|Sep 22, 2011||FPAY||Fee payment|
Year of fee payment: 4
|Jan 15, 2016||REMI||Maintenance fee reminder mailed|
|Jun 3, 2016||LAPS||Lapse for failure to pay maintenance fees|
|Jul 26, 2016||FP||Expired due to failure to pay maintenance fee|
Effective date: 20160603